Oral Abstract Session
SCMR 22nd Annual Scientific Sessions
Background: Restrictive cardiomyopathy (RCM) is composed of disorders with common physiology, but different etiologies, half of which may not be identified [1-3]. The overall prognosis of RCM is poor with progression to heart failure and increased mortality [1-3]. The primary aim of this study is to evaluate the association between cardiac magnetic resonance (CMR) imaging variables (ventricular and left atrial strain imaging) and major adverse cardiac events such as death. The secondary aim is to assess the difference of CMR imaging variables in RCM between women and men.
Methods: 98 patients with RCM (30 women and 68 men); age 61 ± 13 years referred to CMR from 2007 to 2015 were included in the study. All patients were followed to date to evaluate major adverse cardiac event, diseased status. The CMR exam consisted of: 1) cine images (horizontal, vertical, and left ventricular outflow tract long axes and a stack of parallel short-axis slices covering the heart), which were used for measuring indexed left ventricular (LV) mass, ventricular volume, ejection fraction (EF), and ventricular and left atrial (LA) strain (Figure 1 ); and 2) late gadolinium enhancement (LGE) images, which were used for assessing myocardial scar. Univariate and multivariate logistic regression analyses with stepwise elimination of variables were performed to identify CMR variables associated with mortality. The statistical model included all CMR imaging variables and cardiovascular disease risk factors. p< 0.05 was considered significant.
50 patients (51%) had multiple myeloma and 39 (40%) had amyloidosis. 46 (47%) patients demonstrated signal enhancement on the LGE images and 35 (36%) experienced death. While mortality in patients with RCM was significantly associated with lower body mass index (p=0.03) and higher indexed LV mass (p=0.05), only LA longitudinal strain (p= 0.001) (Figure 2), older age (p=0.05), and presence of amyloid and diabetes (p=0.02 and 0.05) remained significantly associated after adjustment in the regression analysis (table 1). The presence of LGE was not associated with mortality (p=0.29). Women with RCM demonstrated higher LV circumferential strain peak (mean ± SD = 7.4 ± 1.7 versus 6.5 ± 1.6, p=0.03) smaller indexed ventricular end-diastolic volumes (p=0.005 and 0.01 for LV and RV respectively) and smaller LA volume (0.05). Mortality in women was not significantly different than in men (p=0.36).
The results demonstrated that decreased LA longitudinal strain is significantly associated with mortality in patients with RCM independent of other CMR and clinical measures. In conclusion, CMR is a valuable modality for evaluating cardiac strain in RCM patients beyond the need for gadolinium administration.